John Bowen Synth Design

John Bowen wrote:I wonder if Cubase treats USB MIDI differently than via the MIDI interface you have?

Nope. I tried using an alesis q49 controller via usb - connected directly to my pc. Cubase recognized it, and it worked perfect (although, obviously, it's a controller. not a synth which also needs to be controlled)

Hmmm....well, I'm able to connect via USB, sending and receiving MIDI - I'm using Logic Express on my MacBook. Sorry I don't know cubase to know where to check.

John, if you are connecting your unit via midi in ( and not USB ) do you have the same problems i have, when i connect it to live or to Korg Kronos ( you have some examples before in this thread ) ?
It's annoying, because in some configurations i can't use USB ( because midi thru is not in function with USB, wich is normal ) and one of my soft cannot address midi out and USB in the same time. ( The soft FINALE, for example ) And some of my synth didn't have a midi USB port.

I've heard your demo file, but so far, I can't repeat it. I'll keep checking...

I have done some more experiments :

My DAW is ableton live 9 : i have some problems with midi and Solaris as you can see before.

I try to reproduce the bug with Finale : no way, it seems ok.

Of course, ableton correctly drive all my others Synth : Korg Kronos, V-Synth, SY-77, Doepfer Modular interface, Tempest.

If i play the solaris thru the midi out of Korg Kronos : another type of bug, wich i haven't with my other gear.

I try to reproduce the bug with a master Keyboard ( CME UF 80 ) : no way, all seems OK.

So the midi's Solaris seems to uncorrectly interpret the midi flow only with some other materials : Live Ableton ( edirol UA101 midi interface, the same interface used by Finale ), Korg Kronos. And that only if Midi in is used, USB is OK with Live.

Yes, thanks! This is MOST interesting!

As John knows I've been battling this problem for sometime in conjunction with Cubase.
I've tried 5 pin midi & USB. The problem exists on both protocols - but for me is worse with USB.
I'm convinced this is all some kind of NRPN feedback loop.

Neil wrote:As John knows I've been battling this problem for sometime in conjunction with Cubase.
I've tried 5 pin midi & USB. The problem exists on both protocols - but for me is worse with USB.
I'm convinced this is all some kind of NRPN feedback loop.

I'm not sure of that, because i've midi problem with only midi in plug.

I've now got similar problems. Whenever I send MIDI clock to the Solaris it causes stuck notes.

I'm using the 5 pin DIN ports recording and playing back through a MOTU MIDI Express XT from Logic Pro X. I've tried sending a clock signal from Logic and from a Sync Lock separately and both cause stuck notes. As soon as I stop sending MIDI clock everything returns to normal (after I clear the stuck notes).

Hmmmm, my guess is the MIDI buffer may not be large enough. We are waiting for the coder to return from a long vacation, so I hope to have some response on this topic soon.

That would make total sense with my experience of trying to use it with midi from the Cirklon. It sends out quite a lot of data regarding song position etc and note ons and offs are sometimes missed. A small buffer would certainly explain it! Seems the holiday buffer is larger than the midi buffer

I ran into this with program change messages and Ableton. I used a MIDI scope and the thing that seemed amiss is that the Solaris seems to ALWAYS echo program change control messages. So, if you have a MIDI track in record or monitor mode and initiate a program change from either Ableton OR from the Solaris, the program change gets sent to the Solaris, echoed, and this repeats forever. (Unplug a MIDI cable to regain control of your Solaris.)

I'm not 100% sure... but I don't think the Solaris should ever ECHO a program change message. It should respond to program change messages it receives. And it should send program change messages when there's a program change initiated from the synth, and that's it. (O, and it should probably honor the LocalOff setting.)

The workaround I'm using is this: don't set program change messages in Ableton, and don't change programs while the Ableton MIDI track is in record enable mode.

I'm using Ableton Live 9, MIDI DIN ports (not USB MIDI), and the interface on an RME Fireface UFX.

Yes, this is a reported bug in the OS. I think it came from a misunderstanding - and should be a simple fix.

Cool. It's nice to have the confirmation...

This MIDI loop bug with Program Select has been fixed.